Precise voltages from imprecise parts..

Baer

to get even long ones - 1.25 inch long. The quality of that that kind of tr im-pot does depend on the manufacturer, the nature of the linear resistive element and the construction of the moving contact - Vishay's top-of-the-li ne trim-pots had a multi-leaf contact, which gave very smooth adjustment, a nd no no-contact positions along the track. But even cheap and nasty 19mm t rimpots were easier to set than 3/8" multi-turn trim-pots.

I'm bashing on open right now.

here.

Interesting, any idea what the motion does? Clean the wiper?

George H

"

com

com

What about an oscilloscope would care about a fraction of a tenth of a per cent?

If a 10 cm drop decalibrates a scope, it can never be in cal.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

My understanding is that it "seats" the wiper. Same idea as when you get new pads and rotors on your car you want to get good area contact between the two so it takes a bit of spirited braking to get rid of crud on the two surfaces and get the two in intimate contact.

However it works, it was proven empirically to be better (with the T**** pots anyway, but all cermet pots are of very similar construction).

--Spehro Pefhany

On a sunny day (Tue, 21 Apr 2015 07:17:28 -0700) it happened John Larkin wrote in :

Some of those were so difficult to adjust, just 'tick' a fraction of a little of almost nothing. You trust those signal teefees far too much I think.

Does not look to good for your specs... :-)

That sounds like a design problem. A multiturn pot wouldn't help. I have to keep reminding my test techs that

A. Engineers are not perfect.

B. If something is hard to test or calibrate, stop trying and make engineering fix it.

Multiturns store mechanical energy in ways that singles don't. That can make them more sensitive to shock. But no design should depend on setting a trimpot to better than 0.2% of its range, preferably 0.5%.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Some products don't have a CPU! And it gets tricky or expensive to do software-driven gain trims at a GHz. Now and then, a trimpot is ideal.

The nice thing about a trimpot is that all you need to adjust it is a screwdriver. No USB dongles, no software.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

What? No. I think I'm going into shock. What's that pain in my chest.....

On a sunny day (Tue, 21 Apr 2015 09:03:52 -0700) it happened John Larkin wrote in :

We had that design group come over from 'merrica one day. One question I asked is: "Why no anti aliasing filter on the input of a digital scope?"

You proably know much of what you see on teefee is fake, or set in scene if you will, to the point of selling hamburger stands as mobile labs of weapons of mass destruction.

So I never got a straith answer I think.

Of couse, once you think of it, it is 'sjeeper'.

But what you see is most certainly not what you got ?

That is why I love my Ana Log scope so much (well not Anna, but that is an other story altogether).

At least it has a correclation with reality on an understandable level.

I can tell you that 10 turn pots, gears in tuning mechanisms were in every radio, look ahere:

We shoot not General Ice so to speak. man what a language.

Singles live it out on facebook.

That is all true. I have sinned, use a course and than a fine trimpot. :-)

Sorry.

Just uploaded some pics of the inside of two trim pots.

and here,

Both look equally susceptible to back lash in the worm drive.

George H.

When you make that final fractional turn to get the frequency or whatever just right, you leave force wound up in the system. Later, some shock or temperature change will relieve the force, and either the wiper or the screw shaft will move.

There are no precision bearings in sight! It's all plastic.

--

John Larkin         Highland Technology, Inc 
picosecond timing   laser drivers and controllers 

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

Exactly. Why the backlash matters (other than just making the person setting them crabby). One can just tweak it back a bit and relieve the pressure iff the circuit sensitivity isn't insanely high. I actually do (did, I don't do it much these days) that on the multiturns, otherwise they would move a hair if you tapped them.

Nah, no precision bearings. Some rotary ones have the wiper crimped to metal in the center of the alumina so it forms a bearing of sorts. At least radial differential CTE doesn't change the wiper position.

Single or multi, they're settable to maybe 0.1% but really ~1%-ish devices- if you use them carefully (in the best circuits) and if you care about the long term, in a moderately variable environment. If you don't they might be worse (for example, very low resistance pots used as rheostats).

--Spehro Pefhany, Trexon Inc.

[about shock sensitivity of as-trimmed trim potentiometers}

An old treatment for this, was to apply a vibratool (remember the old Burgess scribes for metal marking?) while adjusting. A more modern treatment, would be to have a robotic screwdriver (typical for this kind of adjustment) that oscillates around the final setpoint before stopping, to measure the backlash and by turns counting, compensate it.

Speaking of precise from imprecise, I just got an HP Harrison 6111A Precision Power Supply. (20 V, 1 A, with digital thumbwheels).

It looks to have been made in the late 1960s, and once I ran the swithches through their ranges a couple of times to clean them off, it's within 0.05% over the full range at no load. (There's a few mV of offset.)

Not bad for an all-discrete design with 5% resistors!

Cheers

Phil Hobbs

UPDATE:

1) Never follow the instructions - Refer to the TI LM185-ADJ, LM285-ADJ, LM385-ADJ data sheet page, figure 19 (25V regulator). a) 120K not common (40 of over 100 thousand resistors), 2.4M not common (12 of over 100 thousand resistors); actual output voltage, if used specified values, would be 26.04V nominal; crappy. b) Use of 124K and 2.37M is better.. get 24.9V approximately and easy to trim to better than 0.1% (so far 1K to 12K in series with 2.37Meg resistor). 2) Never use what you get without actual measuring the parts yourself. Got, from DigiKey, continuous tape of 25 each CMF552M3700FEHB resistors and so far randomly found 4.32Meg resistors side-by-side the 2.37Meg resistors; they were labelled correctly. But the fact that incorrect values were on the same strip beggars the question of separating by looks (ie: labeling) alone; incorrect values but same marking could have been there.

2.4M and 120K are standard E24 values. Back when the datasheet for that part was written 1% resistors were a rare and valuable commodity- we traded many a baer skin and stone knife for a reel. Not everyone can afford $1.50 for a resistor.

That's a bit disturbing. I assume you actually meant CMF552M3700FEHB. But the parts were correctly *marked*?

That's what you get for buying those fancy boutique parts. Nobody can afford to make different parts in a reel of 5,000 0603 resistors different values (not at 1/1000 the price).

--Spehro Pefhany

WOW!!

See my posting on abse.

The resistors on the thumbwheel cards are probably better than that, of course. ;)

I love those HP 61xx supplies. I used to have the 3 kV one (6110A), but it made sizzling sounds from the transformer above about 1100V, so I tossed it.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

Sometimes, soaking the transformer in a bucket of thin varnish for a week will fix that.